The Emergent Goebbels: Surveillance Capitalism and the Empty Room

Joseph Goebbels sent a student to transcribe Edward Bernays’ college lectures. Eighty years later, Silicon Valley rediscovered the same methods by accident, optimizing for engagement. The democratic failure of the Iran war didn’t begin in Washington. It began in a classroom.

I. The Student in the Lecture Hall

Picture this: late 1920s, maybe early 1930s, a man slips into Edward Bernays’ NYU class, notebook in hand. He’s not there to earn a degree. He’s there on assignment from Joseph Goebbels.

Fast forward to the aftermath of the war: Allied forces comb through Goebbels’ library and turn up not just Bernays’ books, but meticulous transcripts of his lectures — some predating their official publication. Goebbels didn’t just stumble across these by chance. The simplest explanation is also the most unsettling: he sent someone to collect the source material straight from the tap.

When Bernays found out after the war, he was, understandably, rattled. It wasn’t the political use of his techniques that got under his skin — he’d always argued that shaping public opinion was a civic duty, best left to the experts. What unsettled him was the idea that someone had quietly sat in his classroom, not to learn, but to weaponize his ideas for ends he never intended. He built the machinery of influence to serve democracy. Someone else saw how easily it could be repurposed.

This deserves a closer look. Goebbels wasn’t just leafing through a popular book — he dispatched someone to secure the entire playbook before Bernays had even finished writing it. That’s not idle curiosity. That’s a calculated move by someone who recognized the value and made sure to lock it down. The transcripts aren’t just evidence of influence; they’re proof of a methodical effort to capture the entire system.

But the real story doesn’t end with Goebbels or even Bernays. Fast forward eighty years, and Silicon Valley engineers rediscover the same principle — no need for spies in lecture halls this time. Instead, they ran live experiments on billions, all in pursuit of what keeps us glued to our screens. On one side, you have meticulous planning; on the other, a happy accident. The result? Pretty much identical. That’s hardly comforting, but it’s exactly the point.

II. The Operating Principle

Goebbels caught onto a lesson from Bernays’ lectures that Silicon Valley would rediscover much later: you don’t need to dictate thoughts if you can shape the setting where those thoughts form. That’s the thread we’ve been following since The Persuaders.

​Bernays staged his lesson on Fifth Avenue in 1929. The women lighting up during the Easter parade weren’t acting on a whim. The crowd, the press, the buzz about women’s rights — it was all orchestrated. Suddenly, smoking looked like freedom before anyone paused to question it. The feeling arrived first, the logic trailed behind. The trick went unnoticed, but the product took the spotlight.

​This trick works because of how our brains are built, not because of any clever gadgetry. If something is right in front of us, it feels real. If it matches what we already believe, we’re more likely to accept it. If it looks like business as usual, it feels natural rather than staged. These aren’t personal flaws; they’re just the way people operate. The real move isn’t to beat reason, but to load the dice before reason even gets a chance. By the time we start thinking it over, the outcome is already leaning one way.

​Goebbels took this insight and scaled it up with state power. The cheap radios distributed to German households became the delivery system. Repeat the message, dial up the emotion, make people feel both threatened and included — classic Bernays, only with a darker edge if you tried to tune out. Goebbels’ real leap wasn’t the method, but the scale: the environment had to cover everyone. Leave anyone outside, and the illusion cracks. The more total the environment, the stronger the effect. A platform that reaches everyone always outperforms one that only reaches most.

​Silicon Valley skipped the history lesson. They just watched the metrics. When Google decided in 2003 to monetize our habits with targeted ads, every major platform jumped in: chase the clicks, track what keeps people hooked, repeat. What actually keeps people engaged isn’t neutral reporting or balanced debate. It’s not the kind of shared reality you’d hope for in a healthy democracy. What really works is content that reassures people they’re right and stirs up their existing fears. Outrage keeps us scrolling far longer than agreement ever could. Division spreads faster than consensus. The algorithm didn’t need to study Bernays; it just watched billions of users and let the numbers speak.

​The Volksempfänger needed state funding and force to reach every household. The smartphone managed it through sheer convenience. Now, the platform isn’t really optional for anyone whose news, friends, shopping, and civic life all run online. Technically, you can opt out, but the cost is losing access to almost everything that counts. Bernays would recognize the setup. Goebbels would recognize the scale. Neither would be surprised by where it all leads.

III. The Polarization Machine

Strip away the buzzwords, and the engagement-polarization equation is brutally simple. Build a platform to maximize engagement in a country with a spectrum of political views, and you’re basically pouring gasoline on the loudest, most extreme voices. Why? Because the outrageous stuff gets clicks, while the reasonable middle gets ignored. No shadowy plot here — just an algorithm faithfully chasing its one goal: keep you scrolling. Does it also supercharge polarization? That’s just the machine running a little too efficiently for anyone’s good.

​Here’s the play-by-play: the confirmation bias flywheel kicks in, just like The Enclosers laid out. Feed people what they already believe, and they’ll stick around, clicking and scrolling. The algorithm notices and doubles down. If you’re a die-hard on one end, your feed becomes a custom echo chamber, amplifying your beliefs if you’re on the other end, same story — just a different soundtrack. The upshot? Two camps, each sealed in its own algorithmic bubble, barely sharing a single fact. They’re not even arguing over the same reality anymore. The platform’s optimization engine has built parallel universes.

​Try debating someone who lives in a different universe. Good luck with that. Building a political coalition? Not happening if you can’t even agree on the basics. Democracy needs at least a patch of common ground to stand on. The platform didn’t set out to dynamite that foundation. It just wanted eyeballs. That the shared information landscape got torn to shreds along the way? That’s collateral damage, courtesy of the engagement machine.

​The mess only deepens when you zoom in on politics. The people who show up for primaries — the loudest, most fired-up, and, not by accident, the most extreme — aren’t your average voter. They’re a different species: the average primary voter, handpicked by the algorithm’s sorting hat. If a politician tries to split the difference and keep most folks happy, they risk getting steamrolled in a primary by someone with the algorithmically supercharged base behind them. It doesn’t matter if the extreme stance is a turnoff for most; what matters is who’s most engaged, and that’s who the algorithm keeps feeding. Compromise turns into a losing bet. Senators who try it get tossed. Senators who want to stick around performed. The algorithm didn’t set out to build this circus, but here we are.

​Cambridge Analytica is where the accidental science experiment morphed into a precision-guided political weapon. The platform had already rediscovered Bernays’ playbook, but Cambridge Analytica took it and ran, this time with a laser focus. They sifted through mountains of data, pinpointed who would react to which emotional triggers, and served up content engineered for one thing: specific, measurable outcomes — don’t vote, vote third party, stay home. It’s Goebbels’ strategy, only with a scalpel instead of a sledgehammer. He had to settle for the broadcast’s blunt force. Cambridge Analytica could target individuals: same principle, just a far sharper instrument.

IV. The Room Nobody Was In

The Iran war handed us a masterclass in how democracy can break down, and not in a way you can shrug off. The U.S. military was running missions in foreign waters, with no real War Powers check, Congress mostly asleep at the wheel, and the public debate that’s supposed to precede war nowhere to be found. The rooms where the adults were meant to be — Senate, oversight committees, public forums — were ghost towns. Out of fifty-three senators, only two bothered to even fake oversight. The rest had vanished.

​So, how did we get here? It’s not that these senators are spineless or cartoon villains. Most have real views on the use of military force. The Constitution spelled out their job. What it didn’t spell out was the political price tag for doing it. Each of those fifty-one senators ran the numbers and saw that silence was the safest bet, thanks to an incentive system engineered over decades. The party base — the loudest, most terminally online — had been spoon-fed pro-war soundbites, and anyone who pushed back got branded a traitor. If a senator wanted to cross the aisle on war powers, they’d be stepping into a political landscape their own voters had been algorithmically filtered out of. So, silence became the rational move for each senator. The irrational outcome for the country: Congress flunked its most basic test — deciding when to go to war.

​Six million people who voted for Joe Biden in 2020 sat out the 2024 election. That’s not a rounding error; that’s enough to flip an election. The SC series argues this isn’t about voters getting more radical, but about people unplugging altogether. When the information machine is tuned to crank out fear and outrage, it also fries the circuits of anyone sharp enough to see the game. The folks who opted out weren’t wrong about the system. They just looked at the chaos and figured their vote wouldn’t count, or worse, that caring would only make them feel worse. The algorithm didn’t set out to sideline the civic-minded. It just chased engagement. The result: the people most invested in democracy were the first to walk away.

​The Iran series mapped what happened from the outside: who was in the room, how the Senate voted, and who stopped showing up altogether. The SC series traced the inside story: how engagement algorithms, confirmation bias, and the ever-present threat machine dissolved any shared reality. Neither told the whole story alone. The Iran series had the receipts, but not the why. The SC series had the why but lacked evidence of its impact on democracy. This article connects the dots. The room was empty because the algorithm made showing up feel pointless, and the political climate made honest debate a nonstarter.

V. The Emergent Goebbels

Goebbels operated with a blueprint in hand: ideology, state power, secret police, and a radio in every living room. He orchestrated the flow of information with clinical precision, fully conscious of the consequences. For him, the machinery of totalitarianism was not just a tool but the entire point — ensuring his methods delivered exactly as designed.

​Silicon Valley took a different route. Here, the algorithm stumbled into Goebbels’ old role by sheer accident, propelled by market incentives rather than malice. Engineers optimized for efficiency, data scientists chased the next uptick, and executives kept their eyes on quarterly reports. No one set out to unravel democracy or fracture the political center; the chaos was simply the residue of a thousand rational business choices, each made in isolation, with the bigger picture left out of frame.

​That’s hardly reassuring. If anything, it’s more unsettling. Goebbels could be prosecuted; intent could be regulated. When someone deliberately trains for subversion, the law has a playbook. But when democracy unravels as a side effect of algorithms chasing engagement, there’s no villain to indict. Emergence doesn’t fit in the dock. The aftermath of a series of perfectly legal, logical business moves slips through the regulatory net.

The rules we’re drafting now target bad actors — those who intend to harm. GDPR is designed for companies eager to misuse data. The EU AI Act is aimed at systems built to stir up trouble. Cambridge Analytica landed in hot water for meddling in elections. But the platforms that enabled all this? They slip through the cracks because their only crime was chasing engagement. The collapse of democratic debate was just an unfortunate byproduct. In this version, the Goebbels figure didn’t even realize what was happening. The result is just as hollow.

​My father fought in World War II. His generation saw up close what happens when Bernays’ playbook is picked up by a regime with state power and a taste for control. After the war, they built an international system — institutions, rules, and guardrails — meant to keep that kind of threat at bay. Those frameworks were designed for a world where governments deliberately sculpt what people know, where a propagandist trains an apprentice and then enforces the lesson with force.

​They never anticipated the accidental version: a platform that quietly colonizes the information landscape, not by pushing ideology, but by optimizing for convenience and engagement. The outcome is eerily familiar, but there’s no mastermind at the helm. They couldn’t have predicted it. The technology didn’t exist. The business model hadn’t been invented. The notion that a private company, simply chasing its next earnings report, might inadvertently dismantle the shared information space on which democracy depends? That possibility never even entered the conversation.

​The student in Bernays’ class scribbled notes for a boss with a plan. The engineers behind engagement algorithms were solving a different puzzle, unaware of the destination. Yet the outcome converged: a landscape where instinct trumps reflection, and the notion of a shared reality — where debate is possible — becomes not just difficult, but nearly unimaginable for those inside it.

​It took a global conflict to halt Goebbels’ version. The platform’s version, by contrast, is woven into daily life for billions — embedded in the phone in your pocket, sustained by the profit engine of the world’s wealthiest industry, and governed by rules that only make sense if you assume intent is the culprit.

​The student’s notes gathered dust in a single library. The platform’s version lives in every pocket. In both cases, the room remained empty.

Cognitive Enclosure Series · May 2026

Part of an ongoing series on surveillance capitalism, autonomous vehicles, and humanoid robotics.

For the mechanism: The Enclosers · From Cigarettes to Cities · Who Writes the Rules?

For the consequences: search “Millennium Challenge 2002” on Medium for the Iran series.

This publication is reader-funded — no ads, no sponsors, no brand deals. Medium’s algorithm pays roughly 25¢ per article and actively buries long-form in favor of clickbait. If you want more writing like this, buy me a coffee. It funds the next one directly.